Аңдатпа
Artificial intelligence (AI) has significantly transformed distance learning by enabling adaptive educational environments and large-scale learning analytics. However, existing studies primarily focus on either social interaction or academic outcomes, while the integration of content-mediated, system-supported, and behavioral interactions remains insufficiently explored. This study aimed to identify learner profiles and interaction patterns using educational analytics data from an AI-supported distance learning platform. The study analyzed interaction logs from 721 university students enrolled in an intelligent distance learning platform. Clickstream data, learning resource transitions, and behavioral actions were examined using learning analytics, network analysis, k-means clustering, and analysis of variance (ANOVA). An immersion index was introduced to evaluate activity accumulation and attention retention across educational resources. Five learner profiles were identified: Observers (77.53%), Evaluators, Universal learners, Commentators, and Content Creators. The findings demonstrated statistically significant effects of learner profile, resource type, and their interaction on engagement dynamics ( p < 0.001). Universal learners exhibited the most interconnected interaction networks, whereas Observers showed broad but low-density interaction patterns. Resource preferences differed substantially across learner profiles, indicating heterogeneous participation strategies. The findings demonstrate that meaningful engagement in AI-supported distance learning extends beyond visible social participation and includes content-oriented and observational learning behaviors. The proposed analytical framework provides evidence for adaptive course design, personalized recommendations, and differentiated instructional support based on learner interaction patterns. These results contribute to the development of more inclusive, data-driven, and intelligent distance learning systems.